Khaberni - In a surprising scene that confused thousands of users, the AI-based chat model "Claude" experienced a widespread malfunction on Monday, March 2, 2026.
The malfunction led to significant disruptions in service across various platforms, whether through browsers, mobile applications, or application programming interfaces.
The model, developed by the company Anthropic, faced a partial outage that disrupted the main application, the desktop program, as well as the public API interface, leading to a succession of complaints from users and developers about difficulties accessing or using the service normally.
The Downdetector site, which specializes in tracking digital failures through user reports, showed a sharp increase in the number of reports related to the "Claude" service, including several countries such as Saudi Arabia, Egypt, the United States, and the United Kingdom.
The reported issues varied between failures in loading chat sessions, general error messages when trying to log in, and API responses returning what are known as "500 errors," indicating a server-side problem, not due to the users' devices or their internet connection.
In the first official comment, Anthropic announced through the service status page that they have detected the fault and started investigating its causes. In a subsequent update, they stated that "the cause of the problem has been identified and a fix is being implemented."
The company also mentioned that some API methods are not functioning as they should, confirming that technical teams continue to work on addressing the malfunction, without specifying a clear timeline for the full restoration of the service.
The malfunction did not limit itself to a specific category, as it affected chat users for daily purposes, as well as developers who rely on "Claude" in their applications and technical projects via programming interfaces.
Among the problems also detected, there was an inability to open the desktop application for some users of the "Windows" system, along with expired waiting times and incomplete requests, reinforcing the impression that the malfunction is a widespread internal technical issue.
As of this moment, the exact cause of the outage remains unannounced in detail, while the company confirms that the repair process is ongoing.
